MIVisionX documentation

MIVisionX is a comprehensive machine intelligence and computer vision toolkit. MIVisionX includes AMD’s implementation of Khronos OpenVX, as well as a Convolutional Neural Net model compiler and optimizer supporting the ONNX and Khronos NNEF exchange formats. The MIVisionX toolkit is used for rapid prototyping and deployment of optimized computer vision and machine learning inference workloads on a wide range of computer hardware, including on small embedded CPUs, APUs, discrete GPUs, and heterogeneous servers.

The MIVisionX public repository is located at ROCm/MIVisionX.

Installation

• MIVisionX prerequisites

• Build and install MIVisionX for Linux

• Install MIVisionX using the package installer on Linux

• Install MIVisionX on Windows

• Build and install MIVisionX on macOS

• Install the model compiler on Linux

• Install AMD OpenVX

• Test the MIVisionX installation

How to

• Use the model compiler

• Use MIVisionX with Docker

Reference

• AMD OpenVX

• MIVisionX toolkits

• MIVisionX supported models and operators

• MIVisionX environment variables

• MIVisionX API reference

Examples

• MIVisionX sample applications

• Model compiler samples

To contribute to the documentation, refer to Contributing to ROCm.

You can find licensing information on the Licensing page.